Forward Deployed Infrastructure Engineer - APAC
Hyperbolic is an open-access AI cloud providing on-demand GPU infrastructure, reserved and private cloud capacity, managed inference, and AI compute services.
Funding history
About Hyperbolic
Hyperbolic operates an AI cloud platform for training, fine-tuning, and serving AI models at scale. It provides self-serve GPU instances and clusters, reserved capacity, private cloud infrastructure, OpenAI-compatible inference APIs, dedicated model hosting, and related infrastructure services for developers, researchers, startups, and enterprises.
Skills
About the Role
You will be the engineer embedded with reserved customers running large multinode GPU clusters. You stand up and hand over clusters, own customer environments rather than tickets, and take on deep escalations from support engineers. You will also liaise with hardware providers, bringing evidence of faults and coordinating fixes, while automating repeat work and building monitoring.
Requirements
- Deep Linux experience and comfort in the CLI
- Hands-on multinode GPU experience: NCCL, InfiniBand or RoCE, collective performance debugging
- Hardware fault triage on GPU nodes: XID and ECC errors, dmesg, lspci, nvidia-smi
- Experience provisioning and operating GPU clusters with Kubernetes, Slurm, or both
- Comfortable customer-facing
- Sound judgment on when to dig and when to escalate
Responsibilities
- Build, validate, and benchmark new customer clusters and hand them over with documentation
- Debug multinode and NCCL performance, GPU and hardware faults, driver and fabric issues, and container and scheduler problems
- Escalate to providers and coordinate maintenance windows, RMAs, and hung nodes with evidence
- Own named customer accounts and learn their workloads
- Own monitoring and alerting with Grafana and Prometheus
- Automate repeat work and turn escalations into runbooks for the L1 tier